The introduction of livestock domestic grazing changed not only the vegetation but also the soil component of the ecosystem (Johnston et al. 1971, Smoliak et al.
Vegetation and soil responses to short-duration ,grazing on fescue grasslands JOIiAN F. DORMAAR,
SYLVESTER
SMOLIAK,
AND WALTER
D. WILLMS
The effects of animal impact on soil chemical and physical properties as well as range condition were measured over a 5-year period to test tbc bypotbesls that animal impact can improve the nutrient and water status of the soil and promote grassland succeadon. A seventeen-pasture short-duration graxing system was establlsbed in 1981 on 972 ba. The pastures were stocked on average wltb 278 cows wltb calves from 1982 to 1986, wbicb was about twice to triple tbe recommended rate of 0.8 AUM/ba. Increased grazing pressure reduced range condition as reflectedby a loss of de&able species such as rougb fescue (Festuccrscubrelkh Torr.). Soil molsture was always bigber in soils of ungrazed exclosures. Soli bulk density increased wblle hydraulic conductivity decreased with grazing. Litter was not signlflcantiy incorporated into the soll with hoof action. Cbltln-N, as a measure of fungrl biomass, decreased significantly under tbe increased grazing pressure. Tbe bypotbesis that animal impact would improve range condition was rejected since impact, in tbe manner applled during tbe study, resulted ln retrogression of the grasslands. Key Words: forage production, soil physical properties,soil cbemical properties, range condition, rougb fescue The primary objective of most grazing management practices is to maximize livestock production per unit area of rangeland while maintaining a sustained forage resource (Heitschmidt and Walker 1983). Short-duration grazing is a system that enables more rigid control of animal distribution with the use of numerous smaller pastures, thus concentrating livestock and permitting timecontrolled grazing. It has been proposed that shortduration grazing will allow conventional stocking rates to be doubled or tripled regardless of range condition at the time of implementation (Savory 1983). Improved carrying capacity is, presumably, obtained through a positive impact of animal activity on water, nutrient, and energy cycles and, ultimately, increased forage production through advanced plant succession. Conversely, the hypothesis predicts range deterioration in the absence of animal impact. Soil formation may be regarded as a function of, among others, vegetation and organisms (Jenny 1980). Any change in any of the components of the biotic factor may affect the soil. Range soils, as found by the European settlers in the late 19th century, were in relative equilibrium with the existing soil-forming processes. The introduction of livestock domestic grazing changed not only the vegetation but also the soil component of the ecosystem (Johnston et al. 1971, Smoliak et al. 1972). The purpose of this study was to test the hypothesis that animal impact and time-controlled grazing can improve the chemical and physical properties of the soil and promote grassland succession. The objectives were to measure the effects of animal impact on the Authors arc soil scientist, retired range ecologist, and range ecologist, respectively, Aficulturc Canada Research Station, Lethbndgc, Alberta TlJ 4Bl. The authors wuh to thank Mr. and Mrs. Blake Holtman for their coo ration in permitting us access on their ranch and providing us with relevant data. $sprojcctcouldnothave succeeded without their help. Mr. R.G. G&aid and B.W. Keslerassisted with species sam ling while various summer assistants clipped forage plots. TEc study was funded in part by a grant from the Farming for the Future Council of Alberta Agriculture. Manuscript accepted 31 October 1988.
252
soil as well as range condition over a 5-year period. Site Description The study was conducted on the Shipwheel Ranch northwest of Fort Macleod on the edge of the Porcupine Hills (49047’N Lat, 113O39W Long), Alberta, Canada. The plant communities represent the interface of the Rough Fescue Grassland and the Mixed Prairie with the former being on the upper slopes and the latter at lower,elevations. These grasslands have been described by Moss and Campbell (1947) and their response to grazing was reported by Looman (1969). About 70% of the grassland on the ranch was in fair to poor range condition in 1981 when the grazing system was implemented (Wroe 1984). The recommended stocking rate was about 0.8 animal unit months (AUM) per ha (Wroe et al. 1981). The soils are of the Orthic Black Subgroup of the Chernozemic order (Udic Haploboroll) and developed on Laurentide till overlying sandstone. The Ah horizon consisted of 46% sand and 27% clay, The climate is dry and subhumid, and although the annual precipitation averages about 450 mm, it was 408,322,367,423, and 435 mm from 1982 to 1986, respectively. Materials and Methods A 17-pasture short-duration grazing system was established in 198 1 on 972 ha. The pastures radiated from a central cell which was a single source for water and the location of gates for transfer among pastures. Cattle were put onto the range in early May of each year and, following 3 grazing rotations interspersed with 2 rest periods, were removed in late October. The pastures were stocked with 350,320, 236,235, and 250 cows with calves from 1982 to 1986, which was about twice to triple the recommended stocking rate (Wroe et al. 1981), respectively. Calves were weaned in September. Cattle movement between pastures was timed to prevent heavy use during periods of rapid forage growth and to allow recovery following grazing. Consequently, the grazing periods during rotation 1 averaged about 2.5 days followed by rest periods of about 40 days. As forages senesced, the grazing and rest periods became longer. In rotation 3, the grazing periods averaged about 4.5 days. Five pastures were randomly selected for observations. In 1982 a permanent exclosure, measuring 10 X 30 m, was established in each study pasture at a location sufficiently removed from the center cell to avoid impeding animal movement. Paired observations were made inside and outside the exclosures to compare the differential effects of no grazing vs. intensive short-duration grazing. Sampling outside of the exclosure was outside the zone of potentially excessive fenceline traffic. Basal areas of forage species were determined with a point frame having 35 points spaced at 2.5-cm intervals. In each pasture, 2,100 points were sampled systematically along a transect on both the grazed and protected treatments for a total of 10,500 points on each treatment. Sampling was done in July 1982 and again in June 1986. Percent composition was determined for major species and plant forms. These values were also used to determine range condition (Wroe et al. 198 1) after applying a weighting factor by species
JOURNAL
OF RANGE
MANAGEMENT
42(3),
May 1989
(Lodge and Campbell 1965) to convert basal area to a dry weight basis. The Ah soil horizon was sampled, in 3 subplots each paired within and adjacent to each of the 5 exclosures, on 7 May and 15 Oct. 1985, and on 30 Apr. and 30 Sep. 1986. Samples were handsieved through a 2-mm screen, and stored in sealed, double polyethylene bags at 4O C. At the time of sieving, roots and other debris were removed from the soil and discarded. Moisture content of the soil was determined gravimetrically. Enzymes accumulated in soil have biological significance (Dormaar et al. 1984) as they participate in the cycling of elements and thus play a very important role in the initial phases of the decomposition of organic residues. Specifically, dehydrogenase activity in soils provides correlating information on the biological activity and microbial populations while phosphatase activity is thought to be directly related to the level of organic phosphorus in the soil. Dehydrogenase and phosphatase activities were determined on the fresh, moist soil within 24 and 48 hours, respectively, after its collection from the field to avoid changes in the activities. Dehydrogenase activity was determined at pH 7.6 by measuring the triphenylformazan (formazan) produced by reduction of 2,3, 5-triphenyltetrazolium chloride when soil was incubated with 2amino-2-(hydroxymethyl)propane-1:3diol buffer(0.5 M)at 30’ C for 5 hours (Ross 1971). Phosphatase activity was determined at pH 6.5 by measuring the p-nitrophenol produced when soil was incubated with buffered sodium p-nitrophenyl phosphate solution (0.115 M) and toluene at 37O C for 1 hour (Tabatabai and Bremner 1969). Following the enzyme analyses, the soils were dried and the subplot samples were combined and mixed. Dry soil colors were rated according to the Munsell(l954) notation. Subsamples were ground to pass a OS-mm sieve. At the time of sampling, undisturbed core samples, 55 mm diam. and 30 mm deep, were taken with a drop-hammer type sampler at O-to 3-cm and 3-to 6-cm depths at each subplot. The core samples were refrigerated at 4’ C until needed. Saturated hydraulic conductivity was determined using the Tempe Cell method (Sommerfeldt et al. 1984). The cores were then ovendried, their mass obtained, and the bulk densities calculated. It has been shown (Johnston et al. 1971, Dormaar et al. 1984, Willms et al. 1988) that grazing affects a number of soil chemical parameters such as total organic carbon (C), total nitrogen (IV), nitrate-N (N&N), and available phosphorus (P). Chitin is a polymer of N-acetyl-2-amine-2deoxyglucose linked in a /3-1,4 sequence. It occurs in fungal cell walls and arthropod exoskeletons (Gould et al. 198 1). Since more permanent aggregation is the result of the development of a network of fungal mycelia (Cheshire 1979), it was hypothesized that chitin values are sensitive enough to be used as a measure of potential aggregation as affected by grazing. Total organic C was determined by dry combustion at 9ooo C for 15 minutes; the evolved and scrubbed COa was collected and weighed. Total N was estimated using a macrokjeldahl procedure, and NOs-N was determined by steam distillation (Bremner 1965). The analysis for available P was carried out according to Olsen et al. (1954). Chitin N was determined and corrections were made for extraction efficiency and ammonium-N as outlined by Gould et al. (1981). The light organic matter fraction was obtained as outlined by Spycher et al. (1983). An increase in the light organic matter fraction could be indicative of litter incorporated into the soil by hoof action. However, nonhumitied materials also contain a more rapidly decomposable fraction and thus a source of available nutrients. Although fungal mycelia aid more permanent aggregation, initial development of structure largely depends on organic matter
JOURNAL
OF RANGE
MANAGEMENT
42(3),
May 1959
which is most readily decomposed (Cheshire 1979). Monosaccharides represent the latter. Hence, the monosaccharide distribution in hydrolysates of the soil samples was determined as outlined by Dormaar (1984) except that the hydrolysis step was modified with the elimination of the 2-hour 72% HsSG4 pretreatment (Dormaar 1987). The quantitative analyses of the alditol acetates were done with a Hewlett Packard gas chromatograph 5840 A. Available forage was sampled around each exclosure. Seven permanent sampling sites in fields A to D and 5 sites in field E were randomly located within the area and single 0.5-m* plots were clipped to determine forage removal on the grazed treatments, at each location before and after each grazing rotation. Utilized forage was calculated as the difference between the first and second harvest. Total utilization over the grazing season was calculated as the difference between total available forage and residual. Total available forage was Hrr + (Hz1 - HE) + (Hsr - HZ) where Hi is the harvest at rotation i (1,2, or 3) and time j before (1) or after (2) grazing. The data were subjected to an analysis of variance using a split plot model. The fields were regarded as replicates, the grazing treatment as the whole plot, and the repeat measurements (over seasons, years) on each whole plot as subplot treatments. For the soils data, analyses were carried out for each variable for each year separately and over the years. Since the whole plot error (Error A) was generally smaller than the subplot error (Error B), another analysis was carried out with these errors pooled. The analyses of the hydraulic conductivity at depths 1 and 2 and the vegetation data, including estimates of range condition, were carried out on log-transformed data to meet assumptions of normal distribution (Steel and Torrie 1980).
Results Forage production and utilization averaged 509 and 425 kg/ ha, respectively, from 1983 to 1986 (Table 1). Percent utilization of Table 1. Average stocking rates over 5 fklds md anoual production ud utlllution (if 1SD) in the vicinity of the study sltea from 1983 to 1986. 1983
1984
1985
1986
3.0
2.3
2.5
2.7
1983-86
stocking rates (AUWha) Forage production
(kg/ha) Residual (kg/ ha) Utilization: Actual (kg/ ha) Proportion (%) Total Rotation 1 Rotation 2 Rotation 3
2.6
570 (338) 337 (186) 347 (137) 784 (451) 509 (353) 90 (88) 97 (52) 58 (34) 94(81) 85 (61) 481 (336) 240 (170) 228 (122) 690 (424) 425 (338) 82(16) 26(32) 52 (27) 61(26)
67 (18) 34 (25) 45 (28) 41(24)
83 (8) 14 (20) 56 (20) 46(28)
87 (10) 13 (25) 17 (22) 82 (14)
80 (15) 22 (26) 42 (28) 58(28)
available forage, over the same period, averaged 22,42, and 58 for rotations, 1, 2, and 3, respectively. It is clear from the data in Table 2 and their statistical treatment by treatment and season within and between years in Table 3 that, in spite of sometimes significant interactions among treatments (i.e., grazed vs. ungrazed range) and time of sampling within years or among treatment and time and year of sampling, the grazing management practiced over a 5-year period on fescue grassland had a most pronounced effect on many physical and chemical properties. Physical Properk In 1986 the color of the soil of the exclosures was black (5YR
2/ 1, dry); around the exclosures it was dark reddish brown (5YR 2/2-3/2, dry). This suggests either a loss of organic matter or 253
T8blc 2. Pbyehl8nd chemhl properth of Bhck Chernozcmic soil uoder skortdumtlon gr88lng (Sag) 8nd uwr8zed (Excl) Fescue Crawland 8t Fort M8ckod, Albert8 (8verage Of 5 88mpke). 1985
1986
May 7
Oct. 15 Excl
Sdg
20.2 0.81 0.98
Sdg 27.5 0.94 1.02
29.2 0.82 0.99
13.4 0.92 1.00
4.61 3.29 921 89 4.21 13.6 0.39 11.0 10.2 0.55 9.35 5.04
6.29 4.50 1144 138 4.52 16.8 0.43 10.5 12.6 0.61 7.00 7.66
1.96 2.95 708 177 3.51 14.9 0.39 8.9 10.3 0.64 5.90 5.46
4.51 2.54 913 242 4.23 17.0 0.42 9.1 8.4 0.78 7.13 6.45
0.58
0.86
0.60
0.56
Excl
Sdg 16.1 0.97 1.09
Moisture (%) Bulk density: O-3cm (Mg/ rns) 3-6 cm Hydraulic conductivity: O-3cm (cm/ hour) 3-6 cm Phosphatasc activity’ Dehydrogenase activity* Carbon (%) Light fraction C as % of total C Nitrogen (%) C:N ratio NOa-N oCg/g) Chitin (mg N/g soil) Available P bg/g) Monosaccharides (g/kg)
Apr. 30
galactose + mannose arabinose + xylose
Sep. 30 Excl
Sdg
Excl
17.1 0.84 0.96
22.7 0.94 1.04
28.0 0.83 0.97
4.31 3.15 904 81 4.02 11.4 0.42 9.6 5.6 0.58 4.95 3.89
5.34 4.24 1084 114 4.52 20.8 0.46 9.7 10.0 0.66 4.02 1.07
1.79 2.66 738 186 4.27 12.7 0.41 10.4 8.3 0.71 4.32 5.06
4.69 3.12 879 264 4.90 15.0 0.45 10.8 5.9 0.81 4.98 5.10
0.60
0.78
0.14
0.68
nlcascd, rg/g dry soil/hour. IP-nitrophenol 2Formazan released, nmol/g dry soil/hour.
differential rates of organic matter accumulation and/or decomposition between grazed and rested treatments. Moisture was always significantly higher in the soils of the exclosures. Bulk densities did not vary over the seasons or years, but increased significantly with grazing. Hydraulic conductivity at the O-3 cm depth was negatively affected by the grazing regime.
Enzyme Activities Phosphatase activity decreased over the summer and returned to the previous year levels the following spring. Conversely, dehydrogenase activity doubled from spring to autumn. T8ble 3. Slg8lfiunt effects 8t EO.05 (+) 8nd EO.01 (+*) of physkal8nd cbemhl cb8r8cterleiia of Blrck Chernozemlc eoll8fter ahortduratlon gr8zing (Sag) 8nd Mmpled 8t different 8e8som @ems) 8nd ycrrs (Ye8r). 1985 Sdg ++ Moisture (%) Bulk density: O-3cm *+ (Mg/m3) 3-6 cm l Hydraulic conductivity: O-3cm ** (cm/ hour) 3-6 cm Phosphatase activity’ l* Dehydrogenase activity*** l* Carbon (%) Light fraction C as % +* of total C ** Nitrogen (%) C:N ratio NOs-N olgl g) Chitin (mg N/g soil) ++ Available P ho/g) l Monosaccharides ** (g/kg) galaCtOse+ nlaMOSe l
1985 + 1986
Sdg
Seas
Sdg
Seas
Year
l
l
*+
**
*
l
++
St
l*
l*
l*
l*
*+3
**
**3
l*
**
*+
l*
l*
**
l*
l*
**
+*
l*
*+
l*
.+
l+
l
**
l*
l*
l*
**3
* ** * l S. +*
**
l*
l*
-3
l
** c3
IP-nitrophenol released, rg/g dry soil/hour. 2Formazan released, nmol/g dry soil/hour. YSdgX seas interaction sigruficant. Seas X year interaction mgnikant.
3
**
394 4 4 * 4 l* 3 l* l * 3,4 l
**3 l*
3 3
** l
**
**
arabinose + xylose
264
1986
Seas
** l*
*+ 3
Chernied Properties The grazing treatment showed significantly lower C and N contents of the soil. There was a significant difference over the season for the C content as well. Conversely, the N content remained the same within the years with only a slight shift over winter. The light fraction C as percent of total C represented one form of compartmentalizing the soil organic matter. The autumn samples did have more light fraction C as percent of total C than the spring ones, which could represent increased fine roots over the growing season. Litter was not significantly incorporated into the soil by hoof action, regardless of year, as indicated by the light organic matter fraction data. Chitin was measured to give an indication of fungal biomass and thus of aggregation potential. All comparisons were highly significant. Autumn samples had higher chitin levels than spring samples. Nitrate N behaved differently under the grazing regimes over the season. The same was true for available P. Since polysaccharides containing substantial quantities of arabinose and xylose are considered to be of plant origin and those containing galactose and mannose to be predominantly of microbial origin (Oades 1984), the ratio galactose + mannoselarabinose + xylose should be low (
